6 from .common import InfoExtractor
10 compat_urllib_request,
14 from .brightcove import BrightcoveIE
16 class GenericIE(InfoExtractor):
17 IE_DESC = u'Generic downloader that works on some sites'
22 u'url': u'http://www.hodiho.fr/2013/02/regis-plante-sa-jeep.html',
23 u'file': u'13601338388002.mp4',
24 u'md5': u'85b90ccc9d73b4acd9138d3af4c27f89',
26 u"uploader": u"www.hodiho.fr",
27 u"title": u"R\u00e9gis plante sa Jeep"
31 u'url': u'http://www.8tv.cat/8aldia/videos/xavier-sala-i-martin-aquesta-tarda-a-8-al-dia/',
32 u'file': u'2371591881001.mp4',
33 u'md5': u'9e80619e0a94663f0bdc849b4566af19',
34 u'note': u'Test Brightcove downloads and detection in GenericIE',
36 u'title': u'Xavier Sala i Martín: “Un banc que no presta és un banc zombi que no serveix per a res”',
38 u'description': u'md5:a950cc4285c43e44d763d036710cd9cd',
43 def report_download_webpage(self, video_id):
44 """Report webpage download."""
45 if not self._downloader.params.get('test', False):
46 self._downloader.report_warning(u'Falling back on generic information extractor.')
47 super(GenericIE, self).report_download_webpage(video_id)
49 def report_following_redirect(self, new_url):
50 """Report information extraction."""
51 self._downloader.to_screen(u'[redirect] Following redirect to %s' % new_url)
53 def _test_redirect(self, url):
54 """Check if it is a redirect, like url shorteners, in case return the new url."""
55 class HeadRequest(compat_urllib_request.Request):
59 class HEADRedirectHandler(compat_urllib_request.HTTPRedirectHandler):
61 Subclass the HTTPRedirectHandler to make it use our
62 HeadRequest also on the redirected URL
64 def redirect_request(self, req, fp, code, msg, headers, newurl):
65 if code in (301, 302, 303, 307):
66 newurl = newurl.replace(' ', '%20')
67 newheaders = dict((k,v) for k,v in req.headers.items()
68 if k.lower() not in ("content-length", "content-type"))
69 return HeadRequest(newurl,
71 origin_req_host=req.get_origin_req_host(),
74 raise compat_urllib_error.HTTPError(req.get_full_url(), code, msg, headers, fp)
76 class HTTPMethodFallback(compat_urllib_request.BaseHandler):
78 Fallback to GET if HEAD is not allowed (405 HTTP error)
80 def http_error_405(self, req, fp, code, msg, headers):
84 newheaders = dict((k,v) for k,v in req.headers.items()
85 if k.lower() not in ("content-length", "content-type"))
86 return self.parent.open(compat_urllib_request.Request(req.get_full_url(),
88 origin_req_host=req.get_origin_req_host(),
92 opener = compat_urllib_request.OpenerDirector()
93 for handler in [compat_urllib_request.HTTPHandler, compat_urllib_request.HTTPDefaultErrorHandler,
94 HTTPMethodFallback, HEADRedirectHandler,
95 compat_urllib_request.HTTPErrorProcessor, compat_urllib_request.HTTPSHandler]:
96 opener.add_handler(handler())
98 response = opener.open(HeadRequest(url))
100 raise ExtractorError(u'Invalid URL protocol')
101 new_url = response.geturl()
106 self.report_following_redirect(new_url)
109 def _real_extract(self, url):
111 new_url = self._test_redirect(url)
113 return [self.url_result(new_url)]
114 except compat_urllib_error.HTTPError:
115 # This may be a stupid server that doesn't like HEAD, our UA, or so
118 video_id = url.split('/')[-1]
120 webpage = self._download_webpage(url, video_id)
122 # since this is the last-resort InfoExtractor, if
123 # this error is thrown, it'll be thrown here
124 raise ExtractorError(u'Invalid URL: %s' % url)
126 self.report_extraction(video_id)
127 # Look for BrigthCove:
128 m_brightcove = re.search(r'<object.+?class=([\'"]).*?BrightcoveExperience.*?\1.+?</object>', webpage, re.DOTALL)
129 if m_brightcove is not None:
130 self.to_screen(u'Brightcove video detected.')
131 bc_url = BrightcoveIE._build_brighcove_url(m_brightcove.group())
132 return self.url_result(bc_url, 'Brightcove')
134 # Start with something easy: JW Player in SWFObject
135 mobj = re.search(r'flashvars: [\'"](?:.*&)?file=(http[^\'"&]*)', webpage)
137 # Broaden the search a little bit
138 mobj = re.search(r'[^A-Za-z0-9]?(?:file|source)=(http[^\'"&]*)', webpage)
140 # Broaden the search a little bit: JWPlayer JS loader
141 mobj = re.search(r'[^A-Za-z0-9]?file["\']?:\s*["\'](http[^\'"&]*)', webpage)
143 # Try to find twitter cards info
144 mobj = re.search(r'<meta (?:property|name)="twitter:player:stream" (?:content|value)="(.+?)"', webpage)
146 # We look for Open Graph info:
147 # We have to match any number spaces between elements, some sites try to align them (eg.: statigr.am)
148 m_video_type = re.search(r'<meta.*?property="og:video:type".*?content="video/(.*?)"', webpage)
149 # We only look in og:video if the MIME type is a video, don't try if it's a Flash player:
150 if m_video_type is not None:
151 mobj = re.search(r'<meta.*?property="og:video".*?content="(.*?)"', webpage)
154 mobj = re.search(r'<video[^<]*>.*?<source .*?src="([^"]+)"', webpage, flags=re.DOTALL)
156 raise ExtractorError(u'Invalid URL: %s' % url)
158 # It's possible that one of the regexes
159 # matched, but returned an empty group:
160 if mobj.group(1) is None:
161 raise ExtractorError(u'Invalid URL: %s' % url)
163 video_url = compat_urllib_parse.unquote(mobj.group(1))
164 video_id = os.path.basename(video_url)
166 # here's a fun little line of code for you:
167 video_extension = os.path.splitext(video_id)[1][1:]
168 video_id = os.path.splitext(video_id)[0]
170 # it's tempting to parse this further, but you would
171 # have to take into account all the variations like
172 # Video Title - Site Name
173 # Site Name | Video Title
174 # Video Title - Tagline | Site Name
175 # and so on and so forth; it's just not practical
176 video_title = self._html_search_regex(r'<title>(.*)</title>',
177 webpage, u'video title', default=u'video', flags=re.DOTALL)
179 # video uploader is domain name
180 video_uploader = self._search_regex(r'(?:https?://)?([^/]*)/.*',
181 url, u'video uploader')
186 'uploader': video_uploader,
188 'title': video_title,
189 'ext': video_extension,